in blitz I don't care
This is probably one of the hottest topics in the forum of late. Players abandoning losing games (AKA, rage-quitting) seems to be happening on the site a lot, and seems to be happening more than previously. A number of us in the forums have been asking for some kind of response on Lichess's part. What that response should be, if any, is regularly debated.
At least four different approaches to the problem have been discussed.
1. Do nothing.
Steady as she goes. Proponents say rage-quitters are a fact of life of online chess and Lichess is already doing all it should. Just ignore the rage-quitters as best you can. Fix a tea or watch a YouTube video as you wait for the clock to run down.
2. No Escape option.
Create an option where players cannot leave games without immediately forfeiting the game. This would be an option, and not a default feature, because a minority of players have unreliable internet connections and will choose to not turn the feature on. The feature could have an option to allow a total of 30 seconds for a grace period. That could allow for one quick internet hiccup per game. Proponents note that this is a simple, proven solution that's been used by other websites with success, such as ICC. Detractors say that adding another pairing criteria will make it take longer to find games.
3. Detect rage-quitters systematically.
If an analysis of a quit game is made at the time of disconnect, it would be possible to detect rage-quitters and ban them. Without going into too much detail here, it basically works as follows: Every time a player disconnects from a game when he is evaluated at -8 or worse, that's considered a possible instance of rage-quitting (as opposed to an accidental loss of internet connection). If 5 or 10 such -8 disconnects occur in a row (that is, without any disconnects occurring when the player is even or winning), an automatic ban on the account would occur, along with an autogenerated email to the player explaining why action was taken and how to appeal if the player contends the disconnects were innocent. Note that the odds of a player disconnecting from games where he was at -8 or worse ten disconnects in a row are literally less than 1 in a thousand.
Detractors of this method point out that analyzing positions of quit games would increase the load on the server. Proponents argue that the benefits of systematically catching rage-quitters in a way that has no inconveniences to the rest of the community would be worth that additional server load. (That's my position, if you can't tell. If there was ever a justifiable reason for slightly increasing the total server load, this is it.) Further, proponents argue that a systematic solution to the problem of rage-quitters is needed; moderators cannot hope to keep up with the problem by manually doling out punishments to offenders. Other solutions that do not require moderator involvement, such as a No Escape option, will have some effect on the game play of the community as a whole.
4. Increase existing penalties.
Lichess already imposes penalties on rage-quitters manually. The bans start trivially small, on the order of a minute or two, and double in duration each time they are imposed. Some have argued that the penalties should start more severely from the first ban; being sidelined for a couple minutes hardly even qualifies as a slap on the wrist.
Hope this helps.
At least four different approaches to the problem have been discussed.
1. Do nothing.
Steady as she goes. Proponents say rage-quitters are a fact of life of online chess and Lichess is already doing all it should. Just ignore the rage-quitters as best you can. Fix a tea or watch a YouTube video as you wait for the clock to run down.
2. No Escape option.
Create an option where players cannot leave games without immediately forfeiting the game. This would be an option, and not a default feature, because a minority of players have unreliable internet connections and will choose to not turn the feature on. The feature could have an option to allow a total of 30 seconds for a grace period. That could allow for one quick internet hiccup per game. Proponents note that this is a simple, proven solution that's been used by other websites with success, such as ICC. Detractors say that adding another pairing criteria will make it take longer to find games.
3. Detect rage-quitters systematically.
If an analysis of a quit game is made at the time of disconnect, it would be possible to detect rage-quitters and ban them. Without going into too much detail here, it basically works as follows: Every time a player disconnects from a game when he is evaluated at -8 or worse, that's considered a possible instance of rage-quitting (as opposed to an accidental loss of internet connection). If 5 or 10 such -8 disconnects occur in a row (that is, without any disconnects occurring when the player is even or winning), an automatic ban on the account would occur, along with an autogenerated email to the player explaining why action was taken and how to appeal if the player contends the disconnects were innocent. Note that the odds of a player disconnecting from games where he was at -8 or worse ten disconnects in a row are literally less than 1 in a thousand.
Detractors of this method point out that analyzing positions of quit games would increase the load on the server. Proponents argue that the benefits of systematically catching rage-quitters in a way that has no inconveniences to the rest of the community would be worth that additional server load. (That's my position, if you can't tell. If there was ever a justifiable reason for slightly increasing the total server load, this is it.) Further, proponents argue that a systematic solution to the problem of rage-quitters is needed; moderators cannot hope to keep up with the problem by manually doling out punishments to offenders. Other solutions that do not require moderator involvement, such as a No Escape option, will have some effect on the game play of the community as a whole.
4. Increase existing penalties.
Lichess already imposes penalties on rage-quitters manually. The bans start trivially small, on the order of a minute or two, and double in duration each time they are imposed. Some have argued that the penalties should start more severely from the first ban; being sidelined for a couple minutes hardly even qualifies as a slap on the wrist.
Hope this helps.
Hi guys, thanks for the replies. Anyway, it is a great website. I wonder how lichess earn money, as until now I didn’t spend anything here.
I've emailed contact@lichess.org twice about this. Currently waiting on my opponents 18 MINUTES to run out. I'm suggesting an option to manually ping a player once or twice a game, and require a response in 60 seconds or else it's a forfeit. It's so frustrating.
i also think it's a problem. i look for players who've played a lot of games on the sight, then i feel there's a higher likelihood they'll be more uh., reliable. i think i've seen time limits, though, in correspondence that are infinity? is that right? so if someone just let time run out on that - it'd never run out? i must be wrong on this....
KleinerZero Caveman Ideas (in order of brutality)
caveman approach 1: automatically ban all players with a percentage of completed games below 95%
caveman approach 2: filter out above players from the automatic pairing (but allow them to be challenged and challenge other players)
caveman approach 3: if a player disconnects from a game show a popup to that player that will prevent him from playing other games until he either re-joins the game in progress or resigns. "It seems you left a game in progress. What do you want to do? (resign) (re-join)"
measures you can take on your own
- block users who disconnect in suspicious moments
- sometimes you can claim the victory after a while
- start another game in parallel, usually these people don't come back (not sure it's possible to play 2 games at the same time)
caveman approach 1: automatically ban all players with a percentage of completed games below 95%
caveman approach 2: filter out above players from the automatic pairing (but allow them to be challenged and challenge other players)
caveman approach 3: if a player disconnects from a game show a popup to that player that will prevent him from playing other games until he either re-joins the game in progress or resigns. "It seems you left a game in progress. What do you want to do? (resign) (re-join)"
measures you can take on your own
- block users who disconnect in suspicious moments
- sometimes you can claim the victory after a while
- start another game in parallel, usually these people don't come back (not sure it's possible to play 2 games at the same time)
Actually cva 3 is how it is handled on other renowned servers like playchess. But lichess clings to it’s „everyone can do it simultaneously“ til the bitter end.
lichess.org/forum/lichess-feedback/the-root-of-all-evil#1
lichess.org/forum/lichess-feedback/the-root-of-all-evil#1
This topic has been archived and can no longer be replied to.